WTCB SITREP NOV 2022

I wanted to share one of the huge benefits of WTCB. I’ve pretty much completed the WTCB down to 20cM Matches and have added in a number of under-20cM Matches for which I had segment data (from GEDmatch, primarily, and some who had tested at the other companies). These under-20 Matches can be Clustered by looking at their over-20 Shared Matches for a consensus.

There are positives and negatives to WTCB. Overall, a large percentage of the over-20 Matches fit into very solid Clusters. But, just like a distribution curve, some of the Matches do not have many Shared Matches (a few have 0), and some just don’t seem to form a good, solid consensus. If you know me, I focus on what I *can* do – so I want to give you an example of a successful Cluster. And I want to note that this is not the best example, but it is a good one.

Here is a picture of Clusters 54 to 79 in a Super Cluster. The 281 Matches in the Super Cluster range from 20cM to 56cM (the upper threshold was 60cM for this run).

In my review of most of the Clusters and SuperClusters, I’ve found that the individual Clusters look prettier and more solid, but they do not represent a split in ancestral lines within my genealogical time frame (roughly 9 generations back; 8C level). So I have combined most of them into Cluster 54 – a total of 281 Matches.

In this Cluster I now have 3 Matches with an MRCA of A0020 (MITCHELL/UNDERWOOD couple); 12 Matches with an MRCA of A0084 (UNDERWOOD/CANNADAY) and 27 Matches with an MRCA of A0170 (CANNADAY/HILL). I also have 4 Matches who have MRCAs on different lines. The Cluster is very solid, so I suspect these 4 Matches are probably *also* related to me somewhere on my MITCHELL to UNDERWOOD to CANNADAY line. But clearly the 42 Matches on one line show a consensus!

Also within Cluster 54, I have 9 AncestryDNA Matches with segment data – they are all in Triangulated Group [17D25] – another pretty clear consensus. In DNA Painter, I could paint all 281 Matches on Chr 17, from 24 to 45Mbp. Note: In my TG spreadsheet I have over 150 Matches in TG [17D25] – 9 of them from Ancestry Matches and the rest from the other companies.  

I have Ancestors in my Tree beyond A0170 (CANNADY/HILL) which are fairly well known and also in many other Trees, and I’ve found Matches with those more distant MRCAs in other Clusters, but not in this Cluster 54. I’m coming to the conclusion that the 21Mbps in [17D25] probably came to me from either William CANNADAY 1730-1801 (A170) OR his wife, Nancy HILL 1733-1801 (A171).

But the best is yet to come. This Cluster 54 is a classic *pointer*.  I am now pretty sure that the rest of the Matches in this Cluster will have an MRCA with me on the same line. In fact, I’ve only recently found several of the MRCA Matches by building Trees back and/or looking at Unlinked Trees. Here is an example:  

In Cluster 54, I had a 36cM Match with an Unlinked Public Tree with 6 people in it. I opened it up to find only one real lead – Audrey (so I searched Ancestry for her):

BINGO! Note Audrey’s mother is a CANNADAY!! The rest was easy – I quickly found the Match’s link to A0170 (CANNADAY/HILL).

Note: I’ve had others that were just as easy; and some that took more searching and digging; and some that I threw in the towel and moved on.

The bottom line is that the WTCB tool can be very valuable in many cases. And when it works, I’ve got a Cluster which is a great MRCA-focused tool; I’m compiling consensus data for the Cluster (firming the TG and Chromosome Map), adding to the Ancestry Match Notes and helping ThruLines find more MRCAs in Private Trees.

[19Nc] Segment-ology: WTCB SITREP Nov 2022 by Jim Bartlett 20221112

25 thoughts on “WTCB SITREP NOV 2022

  1. Hi Jim
    Your work is impressive. I’m not familiar with your background. I am researching a fully endogamous uncle, so it would be very encouraging to me to know that your methods work with a full complement of endogamy.
    Many thanks,
    Kate

    Like

    • Thanks for the kind words. My background: BSEE 65, MS Engineering Mgmt 81, took an MITx course in DNA; 30 yrs Naval Officer; Construction/Project/Program Management. But the important part is: genealogist since 1974; and into DNA since 2002; tested at all the companies and lecture on DNA. I am not certified in genetics or population science; but I do think I know how atDNA works.
      Segment Triangulation works at the genetic level – segment by segment – each segment coming from a specific Ancestor “box” in your Tree. However Triangulation cannot tell you, on it’s own, which box. Triangulation shines as a grouping tool – every Match in a Group should come from the one line in your Ancestry that passed down the DNA Segment.
      Your comments are on a post about Clustering. Clustering *tends* to group Matches based on a Common Ancestor. It’s strength comes from “solid” Clusters, where all the Matches match almost all of the other Matches. The Clusters form in all different sizes (depends on the Shared Matches), some are large and solid, some are small and more “iffy”.
      Both of these methods are tools – primarily to group Matches and provide “pointers” and clues. Neither is a magic wand that will take the place of genealogy research. Sorry that I cannot give you more encouragement, Jim

      Like

      • Thanks, Jim. By my waffle-y use of “background” I was meaning to ask whether you have had to deal with endogamy to any great extent in your DNA work. Sorry for the miscommunication. I know you have said elsewhere that a segment is a segment. Of course, I cannot argue and I am not whining about the endogamy. Figuring out a suitable strategy that continues to be promising for the long haul is challenging.

        Today’s little surprise was to discover that Ancestry’s new SideView may or may not work for me: A known MATERNAL 2C2R showed up on the PATERNAL side–which, if believed, implies that all 8 segments come from my paternal lines and she has inherited nothing from the side on which I know the MRCA. So… this means either 1) SideView is buggy as heck, or 2) my maternal 2C2R is also strongly connected to my paternal side and I have yet to find out how. I’d hate to spend a lot of time contemplating the latter, if it turned out the truth was the former. Never boring, but frequently frustrating.

        Thanks again for taking the time to write about your experiences. And… I must say your career creds are impressive.

        Kate

        Like

      • Kate, I don’t have classic endogamy, but I do have Colonial Virginia ancestry and a number of my Matches share multiple Common Ancestors with me. I also have 14 known cases within the last 12 generation of cousins marrying each other including one set of 1C marrying. At GEDmatch, however, it says my parents are not related so I’ve dodged some of the problems. I like segment Triangulation because it is a grouping method that is unambiguous. And once done, all new Matches are easily added to extant TGs. So Triangulating (and Clustering) are helpful tools – but not magic wands… Jim

        Like

      • Hi Jim,

        Thank you. I have been working with triangulating groups, as a first step toward some distant day when I can make some sense of it. I am also tree-building, which is challenging because women, particularly, seem to lose their identities quite routinely, for instance… But, it it were any other way, where would be the fun?

        Can you help me by telling me your naming convention for your triangulated groups? 23 chromosomes x 2 sides x n groups… I bet you have already told us this somewhere else in your blog.

        Best,
        Kate

        Like

  2. You made a comment about use gedmatch for ancestry matches. I have matches on ancestry that also took a 23andme test but did not show up. All I had to do was invite them with their email address and they then showed up on my match list. So 23andme limits your match list but this is one way around that. The other is to send out invites to all matches so that they do not count against your limit. I was also on the v5 chip and got an invite for $10 for the subscription to get 5000 matches which is now 5500. I am just mentioning 23andme because there are some that will not use gedmatch. 23andme website was on a list of top security sites also.

    Like

    • RH – another great tip. I’ve been with 23andMe since 2010. And, even then, I sent messages and invites to most – I have 2,500 Matches there. I goofed when the $10 offer was made – I could have doubled my list (however, I have picked up a lot of 23andMe Matches at GEDmatch that are not on my 23andMe list). 23andMe is great for creating Triangulated Groups, but next to 0 for Trees. When I get through most of my WTCB for AncestryDNA Matches, I’ll probably run a DNAGedcom Cluster report for 23andMe to see what kind of corelation I can get with Ancestry. Since I am the base (and, of course, my DNA segments and Ancestors are set), the Clusters should be very close. I know that every one of my Matches who has tested at more than one company, winds up in the same TGs, and within a few rows in my 21,000 row segment spreadsheet.

      Like

      • I was an early tester on v3. I learned about sending out invites. I did for close matches and continued on so that I did all. I go through and send to the new ones once a month. I have over 4000 now. I also tested at v5 and took advantage of the $10 offer. It is interesting my original test will go done to 6cm if they got invites. the invites work if it stays on the match or they accept. if they delete then they can disappear. on the v5 chip with 5000 I still only get down to 20cm. You mention you have colonial Virginia. I have colonial New Amsterdam. I sent out invites. took some time to do all 5k . I have added and an additional 500 since Jan 1. I look forward to reading your assessment of 23andme when you review them.

        Like

  3. Have you used the ability to share match lists on ancestry ? You have to send an invite to do. There are some blogs about doing this. It does help you get past the 20cm and can use dnagedcom. I do not remember reading your blog where you have done that. You can do with all your matches too.

    Like

    • RH, I have not tried that, yet, but want to explore the benefits. I am currently full-time finding MRCAs within most of my Clusters. Most reinforce the Cluster, a few point out issues. Thanks for your feedback. Jim

      Like

  4. Jim, thank you for another great post. Your blog has been so helpful to me over the years as I’ve delved into DNA analysis–the concrete examples and tips based on your personal experience are super and your successes are inspirational!

    My mom and her sister (whose DNA kits are on Ancestry, GEDMatch, and FTDNA) are directly descended from William Cannaday and Nancy Hill (they are 5th great-grandparents), and I was delighted to see this discussion of your CANNADAY/HILL TG. I was hoping one or both of them would be a DNA match with you to add to this group, but alas, it appears you are genetic cousins only. 🙂

    Terri

    Like

  5. Have you ever found any DNA evidence that we are truly descended from the Hill family given the adoption rumors about daughter Misniah Cannaday?

    Like

    • Heather, Please see my other, long winded, reply and let me know what you think. What we really need is an all-female line of descend from Misniah to take an mtDNA test to see if there is any trace of NA. Jim

      Like

    • Actually that is a very good question, and maybe this Clustering can help sort it out. Technically, what the Cluster is saying is that a lot of Matches who descend from William CANNADY 1730-1801 and Nancy HILL 1733-1801 through their other children share DNA segments with me. I have a number of Matches who descend from James N b 1755; William Jr b 1756; John b 1763; and Nancy b 1764 (all reported children of William and Nancy); and I descend from Misniah. To me, that is a strong indication that Misniah was NOT adopted. These are all 6C relationships (or 6C1R or 5C1R, depending on the Match). At AncestryDNA the Shared cMs range from 7cM to 40cM. At this distance, it would be hard to differentiate between full cousins and half cousins – if, say, Misniah was William’s daughter, but not Nancy’s.
      In checking, all together, I have just 4 Matches with descendants of William’s parents (John CANADAY and Margaret SUMMERS, who are A340/341) – this is the 7C level, and I have several other 7C Ancestors with many more Match cousins – at A352 I have 70 Match cousins; at A354 I have 37; at A360 I have 26. So just 4 Matches makes William’s parents suspicious. Or maybe William CANNADAY 1730 is not the father of Misniah (and those 4 Matches are just random because me and my Matches have a lot of Colonial Virginia ancestry). At A342 (Nancy HILL’s parents: John HILL and Mildred GILLIAM) I have 9 Matches who would be 7C. So not a stong case for the potential parents of Misniah, either.
      Each one of our bio-Ancestors has 2 parents, 4 grands, etc. and, depending on the distance back and the size of the families, we should have some DNA cousins with many of them. How many is hard to estimate given all the variables.
      Bottom line: a LOT of Matches with MRCA being Misniah’s parents – it seems like at least one of them should be a bio-parent, if not both. All of those Matches appear to have the same MRCA with me. So it seems Misniah was not adopted. On the other hand, I have very few Matches from the parents of William CANNADAY or Nancy HILL, indicating neither one may be the parent’s of Misniah.
      If you’ve followed all of this, what do you think? Have you found any cousins on Misniah and/or her parents? Jim

      Like

      • I believe over the years I’ve found a little bit of matching on the Cannaday side and none on the Hill side, in the generation of Misniah’s supposed grandparents.

        Like

      • Heather, A few apparent cousins is not a stong indicator of a genetic or genealogy link – particularly with Colonial Virginia ancestry. I have a number of solid genealogy links that are at odds with what the Cluster indicates or what a Triangulated Group indicates. This is why the further back we go, the more reinforcing evidence we need. One share DNA segment plus one Common Ancestor does not equal a solution by itself… you and I have two apparent MRCAs: 5C on UNDERWOOD and 7C on BUSH and one [17D25] TG. From this blogpost, I’d bet on the UNDERWOOD side for [17D25] and no DNA from the BUSH relationship. Jim

        Like

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.